Objectives of this study are to quantify the need for blood transfusion during liver transplantation (LT) and confirm the importance of intraoperative blood transfusion as an independent prognostic factor for postoperative outcome. Furthermore, we try to detect useful variables for the preoperative identification of patients likely to require transfusion of packed red blood cell units (PRCUs) and identify measures to reduce transfusion needs. Data were collected prospectively between September 1998 and November 2000. One hundred twenty-two LTs were included in the study. Forty-two patients (34%) did not require transfusion of PRCUs. In multivariate analysis, transfusion of more than three PRCUs was found to be the only significant variable associated with prolonged hospital stay. In addition, excluding perioperative deaths, PRCU transfusion, using a cutoff value of six units, was the only variable to reach statistical significance (P = .008; risk ratio, 4.93; 95% confidence interval, 15 to 15.9) to predict survival in a multivariate analysis that also included Child's class and United Network for Organ Sharing (UNOS) classification. Moreover, only preoperative hemoglobin (Hb) level was found to significantly predict the need for transfusion of one or more PCRUs. Finally, only UNOS classification and placement of an intraoperative portacaval shunt were found to be statistically significant to predict the need to transfuse more than six PRCUs. We found the requirement of even a moderate number of blood transfusions is associated with longer hospital stay, and transfusion of more than six PRCUs is associated with diminished survival. Preoperative normalization of Hb levels and placement of an intraoperative portacaval shunt can diminish the number of blood transfusions during LT.